Monday, August 20, 2012

Is Amazon Select Free Dead?

Yes! No! It depends!

If all you're using Select Free days for is to bump more sales, then your answer will be shouted from whichever side of the sales curve your book happens to be on after a free run.

And, because things with Amazon are always changing, it appears a waking dragon is starting to make some interesting noise.

Let's look at the trends over the last two months. Oh, wait, there are no trends. Results are all over the map, and each book I'm privy to sales figures for seems to have its own idea of how to perform duing and after its free run. Maybe YOU can spot a trend.

I'm going to arbitrarily break the books down into New Releases, Series, and Standalones because maybe it makes a difference where a book is in its lifecycle. Or maybe not. I'll include genre, free history and site mentions. Note that none of the books I've seen sales figures for has had a paid ad for its free run. For the books I have permission to name, I am. Others I don't, so I won't.

At the end, I'll compare how my strategy today for a free run is different from what I posted way back last February.

And what about that waking dragon, you ask? For US-based authors, it's beginning to look like the UK has raised its sleepy head and is starting to roar. But like the rest of the Select Free equation, if you're on the wrong side of the sales curve, you might not be able to hear it yet. More about that in the individual results below.


NIKO'S STOLEN BRIDE by Lindy Corbin - Contemporary Romance - Standalone

Notes: Definitely an out-of-the-ballpark hit that defies all lessons learned about free runs. New author, no reviews, no major site mentions. Only the smaller sites that auto-list all the free books picked it up.

Free Stats
Free for 5 days, July 12-16 (set free within a few days of publishing)
#5 overall in both the US and the UK
US: 30,000 downloads
UK: 5000 downloads

Sales Stats July 17-31
US: 1800 sales (1460 sales in Week 1 at 99c, 340 sales in Week 2 at $2.99) and 104 borrows
UK: 250 sales (220 sales in Week 1 at 99c, 30 sales in Week 2 at $2.99)

PARANORMAL ROMANCE by a New York Times and USA Today bestselling author - Series

Notes: This book is the beginning of a new series set in the same world as one of the author's traditionally published bestselling series. Unlike Niko above, this book had a "name" author, 6 reviews before it went free, and was picked up by every single major site.

Free Stats
Free for 3 days, Aug 9-11 (set free 2 weeks after release)
#14 overall in the US
~#200 overall in the UK
US: 15,500 downloads
UK: 290 downloads

Sales Stats Aug 12-18
US: 280 sales at $2.99 and 39 borrows (compare to 250 sales at $2.99 during its first 2 weeks in release before going free)
UK: 8 sales

ARROW TO THE HEART by Jennifer Blake - Historical Romance - Standalone

Notes: This is a backlist re-release of an older national bestseller, with high-profile reviews for other editions. It got no major site mentions.

Free Stats
Free for 5 days, July 11-15 (set free within a few days of publishing)
#17 overall in the US
#89 overall in the UK
US: 22,000 downloads
UK: 1000 downloads

Sales Stats July 16-31
US: 340 sales at $2.99 and 58 borrows
UK: 28 sales

SPANISH SERENADE by Jennifer Blake - Historical Romance - Standalone

Notes: This is a backlist re-release of an older national bestseller. It got a Free Kindle Books and Tips mention only from the majors. Mixed reviews since, but we simply didn't see a bump even before the reviews started coming in. Sales are even just a little below pace of the 3 other historical romances released at the same time that did not go free.

Free Stats
Free for 3 days, Aug 9-11 (set free within a few days of publishing)
#33 overall in the US
#179 overall in the UK
US: 10,700 downloads
UK: 430 downloads

Sales Stats July 16-31 
US: 42 sales at $2.99, 8 borrows
UK: 14 sales

DEADLY IMAGE by Tamelia Tumlin - Inspirational Romantic Suspense - Standalone

Notes: With no reviews upon release, we didn't expect this book to get picked up by the majors in July, and it didn't. We did hope for a pickup in August after it got a handful of nice 5-star reviews, but again, no major mentions.

Free Stats
Free for 1 day on July 12, and 3 days Aug 10-12
July: 1250 downloads
Aug: 1050 downloads

Sales Stats
July: 71 sales (60 at 99c, 11 at $2.99)
Aug: 14 sales (10 at $1.99, 4 at $2.99)


Anecdotal only, but based on the collateral sales for other books in a series when one title is free, it does appear that series sales remain strong when free books hit in the Top 100. Several successful books that are not Steel Magnolia titles, but whose authors have kindly shared data, have had collateral sales in the couple of hundred copies during the course of the free run and in the immediate aftermath. Having other books in the series discounted during the free promotion also appears to be quite attractive to readers.


In July and August both, we had books free that had all at one time or another been in the Top 100, many several times and many in the Top 50 once or twice.

Books that received one major site mention landed in the Top 200, but none cracked the Top 100. Each run in July and August, about half the books that were not new releases fell into this group. A far different scenario from having 7 out of 10 free books in the Top 100 in May. I did, however, have 3 books I was managing in the Top 35 in August. So, for a while, we owned 10% of the top third of the Top 100 Free list :o).

Downloads averaged 1000-1500 with a major mention, and 200-1000 without a major mention.

Post-free sales ranged from 30-100 in July and August during the first week post-free for books with 1500 downloads and 10-20 sales for books with fewer than 1000 downloads.


Then there's my thriller, SECTOR C. This book has had a free run each month since January. Data for Standalones above would indicate diminishing returns for books that keep pulsing free. Yet, somehow, SECTOR C keeps plugging along (helped, of course, by the kindness of at least one major freesite mention each run - in July, it got a mention from Ereader News Today and in August from Pixel of Ink and Kindle Fire Department).

A brief history of its free run download numbers:
Jan - 17,000
Feb - 2300
Mar - 2400
Apr - 1900
May - 11,500
Jun - 3850
Jul - 3450
Aug - 9400 US, 1770 UK

Concerned that 1) SECTOR C may have played itself out on the free circuit and 2) total sales each month have been on the downslide, I decided I needed to do something to help inject new life into this little book. Fond as I continue to be about the saber-tooth image the cover originally sported, one thing I'm certain helped this time is the cover change I made. It's much more obvious this is a medical thriller now. And by putting a woman on the cover, I hope to appeal more to women readers.

POI readers seemed to respond well to the new cover; the last time it was featured on POI (in June), it got fewer than 4000 downloads.

Also, it's never gotten more than 200 downloads in the UK before. I think it received so many this time for two reasons:

  1. The new cover has a broader appeal
  2. Freebies in the Mystery/Thriller genre seem to be more popular across the board now than in the past. Another thriller free the same time SC was followed SC up the charts, then went on to dominate overall with an additional 2 free days. And a mystery went gangbusters there, as well. Both authors tell me this is the best their books have done in the UK.

Of interest, I had only 3 days for my free run. UK downloads kicked in hard near the end of Day 2 and the rank was still improving in the UK when I canceled the free run at #17 overall. By then, its US rank had gone from #34 to #55. Another day or two free in the UK would have made a huge difference, I think. I'm not sure it could have rivaled the 5000 Niko's Stolen Bride got in July, but it might have come close.

But a great free run without a sales bump would be disappointing, no? As a pleasant surprise, the UK delivered. Why? Because the UK is still running the old algorithm for its pop lists where freebies are weighted close to the same as a real sale. Great visibility in the UK is the reward for a successful free run.

Sales comparison for the first 7 days post-free for the last 4 runs:
May - 173 sales at $4.29 and 38 borrows
Jun - 142 sales (43 at $4.39, 83 at 99c, 17 at $3.89) and 35 borrows
Jul - 75 sales (45 at $4.39 and 30 at $2.99) and 23 borrows
Aug - 146 sales at $4.39 and 69 borrows in the US, 175 sales at $2.99 in the UK

Current Thoughts on Freebook Strategy

These are editorial in nature only. Typical caveat of "your mileage may vary" applies.

  • New releases (along with virgin titles that have never been free) can still do very well.
  • A free run continues to be part of Steel Magnolia Press' release strategy.
  • We have modified that release strategy to have only 1 title free in each Collection of 2-5 backlist titles of Jennifer Blake's that we're releasing each month. Each of the other books in the month's Collection is getting paid advertising.
  • The new release itself does not need to go free if another in the series can be made free instead.
  • Successful series sell well during a successful free run of one of the titles.
  • Go big. Earlier, it seemed to matter where a book was ranked when it came off free. Now, it's all about the raw number of downloads a book can manage.
  • If the first day's number of downloads is less than 1800, cancel.
  • The algorithms have a built-in 30-day cliff. Each free run is evaluated case-by-case as to the benefit of having a reserve of free days left to always keep that cliff 2 weeks out.
  • We'll likely hold all current books in Select until March, then re-evaluate the benefits then, unless...
  • If things change drastically, we're prepared to make what changes are needed to best optimize our promotional opportunities. 

There's nothing stagnant about this business. We all have to keep trying new things, watching the results, and tweaking our plans on the fly.


Anonymous said...

I agree that it does seem to be about the numbers. Extending to a third day if the numbers are high on day two seems like a good strategy.

R.J. said...

What kind of paid advertising had worked best for you? Where do you advertise?


Sandy Loyd said...

Thanks for the information. Knowledge is power and understanding how the game is always changing helps create better stradegies.

MJ Ware said...

"If the first day's number of downloads is less than 1800, cancel."
That's great advice. Since I write kidlit I think I'll use closer to 1400 or so, but still it's a great way to maximize your free days.

In regards to Select, it's still working great for my kidlit; I'm getting almost 2:1 (2.13:1 to be exact. That's 1 borrow for every 2.13 sales).

As always, thanks for sharing!!

Greg Hamerton said...

Thanks for sharing the figures and insights, Phoenix!

Free runs still work in epic fantasy, to a lesser extent than before.

I've got a 250,000 word series starter at $3.99 (The Riddler's Gift)

Whereas I was getting close to 13,000 downloads in January for a 2-day free run, and a 4%-of-freeloads sales bump (500 'extra' sales) I now get close to 6,000 downloads and a 1% sales bump (60 extra sales) for the combined two weeks post free.

So my free run 'pays' me about $200 (60*$3.99 book less charges and variable royalty rates + extra library loans). It also generates readers for the second book in my series, an impossible conversion to track, but let's call it $100.

It has some added value of boosting my poplist ranking so I can build on it with advertising 'deal days' in the weeks post free, although it hard to get any return from that money and to identify advertising sites that will yield results. I think advertising costs (research time, creating/placing adverts, ad charges) mostly outweigh the boost in sales revenue from raised poplist rank.

Comparing free days to earlier in the year is very difficult because larger downloads caused higher positioning in the poplist which have exponentially greater sales. Also wildly fluctuating sales figures make establishing a base level impossible, so it's hard to analyse how much a sales 'bump' the free days have caused.

Is it worth staying in Select? Free runs are available without being in Select (in a messy way) via price matching, so it's not fair to use the value of free runs as a yardstick. The only measurable value of Select is the Library Loan payments. My book loans an average of 20 copies a month and I get roughly $2 for each loan, so $40.

My experience suggests sales outside of Amazon (iTunes, BN, Smashwords) for fantasy novels were 5-10%, so for me a book that sells on average 100 copies a month would sell at most 10 copies outside of Amazon, earning roughly $25. A more popular book would doubtless get more library loans, so it probably scales up proportionately.

So the money is slightly in favour of staying with Select. With the time saving of managing only one sales channel/format, the ease of timing free day promotions, and the fact you're supporting the company that has pioneered this amazing digital marketplace, I'd say the Amazon Select program still wins.

J S said...

After watching several recent free promotions (Oct/Nov 2012), it appears that Amazon wipes out any rankings of real sales prior to the free day when the next non-free day starts. One title that had 150,000 rank prior, a good free day run up to top 10 in it's category and top 1,000 of all free books showed below a 500,000 rank the next non-free day. It took real sales to recover to positive rankings.

Phoenix Sullivan said...

JS: What happens is that the bestseller lists look on those free days as books not having had any sales for those days. So you're right; you do start over, in effect. That's why small runs can actually be worse than no free runs at all.

Where you get the bounce is on the popularity lists, not the bestseller lists. And that pop list bounce doesn't kick in for about 48 hours...

Anonymous said...

Thanks for the excellent article, Phoenix! It's great to read someone who isn't decrying Amazon Select as a bully's tool. (Maybe I haven't been reading the right blogs.)

My experience with using KS is all over the map too. As you say, this isn't a static business.